DNN library is not found. [Op:Conv2D] when running in PyCharm 关注

Recently I've installed the `waymo` dataset with `pip install waymo-open-dataset-tf-2-6-0 --user` command, and after that the code which used to run stopped running. 
I've noticed that the above command downgraded my `tensorflow` to `2.6` instead the previously installed `2.11`, so after installing the correct version the code runs as previously when I run it in terminal, but in `PyCharm` it fails with the following trackback:

11/Mar/23 13:07:49 - utils.aux_funcs - ERROR - Exception encountered when calling layer 'conv2d' (type Conv2D).

{{function_node __wrapped__Conv2D_device_/job:localhost/replica:0/task:0/device:GPU:0}} DNN library is not found. [Op:Conv2D]

Call arguments received by layer 'conv2d' (type Conv2D):
  • inputs=tf.Tensor(shape=(64, 128, 128, 1), dtype=float32)
Traceback (most recent call last):
  File "/home/USER/projects/QANetV2/qanet/tf_train.py", line 87, in <module>
    trained_model = train_model(
  File "/home/USER/projects/QANetV2/qanet/tensor_flow/utils/tf_utils.py", line 484, in train_model
    model.fit(
  File "/home/USER/anaconda3/envs/qanet/lib/python3.9/site-packages/keras/utils/traceback_utils.py", line 70, in error_handler
    raise e.with_traceback(filtered_tb) from None
  File "/home/USER/projects/QANetV2/qanet/tensor_flow/custom/tf_models.py", line 208, in call
    return self.model(inputs)
tensorflow.python.framework.errors_impl.UnimplementedError: Exception encountered when calling layer 'conv2d' (type Conv2D).

{{function_node __wrapped__Conv2D_device_/job:localhost/replica:0/task:0/device:GPU:0}} DNN library is not found. [Op:Conv2D]

Call arguments received by layer 'conv2d' (type Conv2D):
  • inputs=tf.Tensor(shape=(64, 128, 128, 1), dtype=float32)

I tried creating new `conda` environment, but nothing helps. But still the weirdest thing is that the code runs just fine in the shell but not from `PyCharm`. Does `PyCharm` caches something, because I can't think of other reason for this problem.

Thanks in advance.